Non-insulin dependent diabetes mellitus (NIDDM) affects 11 millions individuals in the United States and is associated with a number of debilitating chronic complications thought to be related to alterations in glucose, lipid and lipoprotein metabolism and hemostasis. These complications signify an enormous cost both physiologically and psychologically on those affected, impacting their quality of life in many ways. Exercise has been shown to improve glucose, lipid and lipoprotein metabolism, hemostasis and quality of life in normal subjects. Exercise has the potential to improve these variables in NIDDM subjects, although this issue has been systematically evaluated in few studies. This study will examine the effects of four weeks of three times weekly aerobic physical training at 65% VO2 max in a group of 15 NIDDM diet stabilized and weight maintained men and women on glucose metabolism, lipid and lipoprotein metabolism, hemostasis and quality of life. A VO2 max determination will be made 48 hours prior to and 48 hours after the four week exercise training period. Indices of lipid and lipoprotein (plasma lipids, apolipoproteins, lipoprotein surface and core lipid metabolism) will be made 72 hours prior to and 72 hours after the four week exercise training period in the fasting resting basal state. Indices of hemostasis metabolism (fibrinogen and fibrinolysis) will be measured 48 hours prior to and after the four week exercise training period in both the fasting resting basal state and immediately post VO2 max determination. Indices of glucose metabolism (serum fructosamine and an intravenous glucose tolerance test) will be made immediately after the initial VO2 max determination and immediately after the final VO2 max determination. Quality of life will be assessed on the same day as the lipid measurements. All measurements will be performed in the morning to control for diurnal variation. The measurements of glucose, lipids and lipoproteins, and hemostasis will be compared pre and post exercise training using a paired T-test while the quality of life data will be analyzed using non-parametric procedure, the McNemar Test.